The purpose of this repository is to process the IR_images obtained from civil infrastructures, analyze and transfer the results to what engineers and decision-makers needed.

Images --> Segment -->Damage density index -->application

---

An new index was proposed to quantify the defects intensity.

DamageDensity: This file foler is to calculate the index "damage density" in a IR image🌳

Framework to calculate the damage density in infrastructure

preprocess.m: Evaluate the performance of different preprocess methods.

postprocess.m: Evaluate the performance of different postprocess methods.

overallprocess.m: The overall procedure to preprocess and segment the IR images. The results includes:

1. all the processed images
2. damages' area

area_account.m: To get the damages' area.

gcul.m: Calculate the gradient on the edge of damages.

---

Results of some lab tests:

It was shown that the index damage density has a positive linear relationship (the correlation coefficient is 0.94) with total volumes of the subsurface voids in the detection area compared to traditional image segmentation methods.

---

---

The image process method:

A demo:

The original image is like this:

IRimage_matlab+IRimage_opencv: These two file folders are mainly used to process and segment images.

Mainly included three main kinds of image processing methods:

• boundary_image.m: segment objects by find the boundary of the object.

• K_means1.m: segment objects by clustering.

• Globalthresholdseg.m: segment objects by global threshold.

2. IRvideo: This file folder is mainly used to process videos🌵

---